How Many Practice Questions Do You Need Before the NP Exam?

There is no official number. Neither AANPCB nor ANCC tells candidates how many practice questions to complete, and no published study ties a count to a pass on the FNP or AGNP exam. What the evidence does support is a method: enough questions to cover every domain on your exam's blueprint more than once, spread across weeks rather than crammed, each one followed by reading the rationale. For most candidates that method produces a total somewhere between 1,500 and 2,500 questions over two to three months, and the method is what matters, not the total.

This article explains why the number is not fixed, what the learning research says about how questions teach, and how to calculate a target that fits your exam, your timeline, and your weak spots.

Why there is no magic number

The two certifying bodies publish the structure of the exam, not a study prescription. Those structures differ, which is the first reason a single question target makes no sense.

Sources: AANPCB candidate handbooks and exam blueprint pages; ANCC certification pages and test content outlines.

The second reason is that candidates do not start from the same place. A provider who has spent clinical rotations managing adult hypertension, diabetes, and COPD needs fewer questions in those areas than a candidate whose placements leaned toward pediatrics or women's health. A count that is generous for one person is thin for another.

The third reason is the most important: questions are not equal units. A question answered quickly, marked right or wrong, and forgotten contributes almost nothing. The same question answered, then studied, then encountered again a week later contributes a great deal. Counting questions as if each one carries the same weight is the error behind most "how many" debates.

What the research says about how questions teach

Practice questions work through retrieval practice: the act of pulling an answer out of memory strengthens the memory more than rereading does. Roediger and Karpicke (2006) showed that students who were tested on material retained more of it a week later than students who restudied it, even though the restudy group felt more confident. The confidence gap is worth remembering during board prep, because rereading a review book feels productive in a way that missing questions does not.

Dunlosky and colleagues (2013) reviewed ten common study techniques and gave their highest utility rating to only two: practice testing and distributed practice. Rereading, highlighting, and summarizing were rated low. Practice testing with corrective feedback produced the strongest learning, which is why the rationale under each question matters as much as the question itself.

Cepeda and colleagues (2006) synthesized hundreds of experiments on spacing and found that the best gap between study sessions grows as the time until the test grows. For an exam eight weeks away, revisiting a topic after several days to a week beats revisiting it the next morning. In practice, this means a topic answered in week two should come back around week three or four, and again in the final review.

Three practical rules follow:

• A question only counts if you read the rationale, including why the wrong options are wrong.
• Questions on the same topic should be separated by days, not stacked in one sitting.
• Mixed blocks (several systems and domains together) belong in the second half of prep, because they force the kind of discrimination the real exam demands.

A working formula for your own target

Instead of adopting someone else's number, derive one in four steps.

1. Write down the blueprint for your exam. For the AANPCB FNP, that is 43 scored items on assessing, 36 on diagnosing, 36 on planning, and 20 on evaluating, with 30 percent of items on older adults and 26 percent on middle adults. For the ANCC FNP outline effective October 30, 2026, the heaviest domain is implementation at 50 of 150 scored items, with planning at 29, assessment at 26, diagnosis at 25, and evaluation at 20.
2. Decide how many full passes through the content you want. Two passes is the floor; three is comfortable for a two-to-three-month plan.
3. Multiply the exam's scored length by a coverage factor. Ten times the scored length is a reasonable middle target: 1,350 questions for the AANPCB exam and 1,500 for the ANCC exam. Fifteen times is thorough.
4. Add a weak-area reserve of 20 to 30 percent, assigned after your first diagnostic block tells you where the gaps are.

These tiers are planning heuristics, not validated thresholds. They assume every question is reviewed. A candidate who completes 2,500 questions without reading rationales has done less useful work than one who completes 1,200 with a careful error log.

Quality signals that matter more than the count

Once the target exists, shift attention to how each question is handled. The following habits separate productive practice from busywork.

• Keep an error log with three columns: what the question tested, why the chosen option was wrong, and the single fact or rule that would have fixed it. Review the log weekly.
• Answer in blocks of 25 to 50 in the early weeks so that reasoning, not endurance, is the limiting factor.
• Read every rationale, including those for correct answers, because a right answer reached for the wrong reason is a wrong answer waiting to happen on a different vignette.
• Tag questions by blueprint domain, not just by body system. Many candidates discover that they are strong at diagnosing but weak at evaluating a treatment response, and only domain tagging reveals that.
• Switch to full-length timed blocks in the final three to four weeks: 150 questions in 3 hours for AANPCB, 175 in 3.5 hours for ANCC. Pacing failures show up only at full length.

Question banks differ in how they report domain performance; the better ones let you filter by blueprint domain and by age group, which mirrors the two tables AANPCB publishes. NPPace's NP bank tags each item that way, as do several other banks, so check the filters before committing to one.

When to stop counting

Readiness shows up in patterns, not in a total. Stop worrying about the count when the following are true:

• Every blueprint domain has been covered at least twice, including the low-weight ones that are easy to skip, such as newborn and infant items on the FNP exam.
• Your scores on mixed, untimed blocks have been stable for two weeks rather than swinging by large margins.
• Repeat mistakes in the error log are disappearing, and new mistakes are mostly in genuinely unfamiliar content rather than in misread stems.
• At least two full-length timed blocks have been completed at the real exam pace with time to spare.

If those are true at 1,400 questions, more volume adds little. If they are not true at 3,000, more volume is not the fix; the problem is in how the questions are being reviewed, and the error log is where to look.

Key points

Related NPPace guides

• How to Read a Primary-Care Vignette Like a Provider: The 5-Step Method
• The Last Week Before Your NP Exam: A Day-by-Day Checklist
• Why NP Exam Questions Feel Different From the NCLEX, and How to Adjust
• FNP Exam Blueprint Explained: What the AANP and ANCC Actually Test